Skip to content

Predicting Bubble Number Density in Magma

^^Patrick Sullivan^1^,  Fabian B. Wadsworth2, Halim Kusumaatmaja3, James E. Gardner4, Tamara L. Carley5, Dork Sahagian6, Edward W. Llewellin1

  • Affiliations: 1Department of Earth Sciences, Durham University, Durham, UK; 2Department of Earth and Environmental Science, Ludwig-Maximilians-Universität München, München, Germany; 3School of Engineering, University of Edinburgh, Edinburgh, UK; 4Department of Earth and Planetary Sciences, Jackson School of Geosciences, The University of Texas at Austin, Austin, USA; 5Department of Geology and Environmental Geosciences, Lafayette College, Easton, USA; 6Department of Earth and Environmental Sciences, Lehigh University, Bethlehem,  USA

  • Presentation type: Talk

  • Presentation time: Monday 16:15 - 16:30, Room S150

  • Programme No: 3.6.7

  • Theme 3 > Session 6

Abstract

Volcanic eruptions are driven by the formation and growth of gas bubbles that exsolve from the melt as it rises to the earth's surface. Bubble formation is favoured by supersaturation, which provides the thermodynamic driver for nucleation, and opposed by surface tension, which gives a free energy cost of to the melt-gas interface. Experimental data has shown that substantial supersaturation is needed to overcome this energy barrier to form bubbles in the bulk melt (i.e. homogeneous nucleation). The presence of crystals in the melt has been shown to reduce the supersaturation needed to form bubbles by reducing the energy barrier to form the phase interface (i.e. heterogeneous nucleation). We present a numerical model for nucleation in a decompressing melt. We combine classical nucleation theory with a probabilistic approach to predict the stochastic formation of bubbles. Heterogeneous nucleation is captured via a reduction in the energy barrier associated with nucleation on a crystal. The model predicts the evolution of the distribution of bubbles in space and time using a Voronoi approach, in which supersaturation at potential nucleation sites is modified by progressive diffusion of volatiles into neighbouring bubbles. The model accurately predicts the evolution of bubble number density observed experimentally in rapidly decompressed, magnetite-bearing melts, across a range of total decompressions. At low decompressions (<20 MPa) we observe incomplete nucleation, with bubbles forming only on some crystals. At moderate decompressions (20-80 MPa), one bubble nucleates on every crystal. At the highest decompressions (>80 MPa) we predict homogeneous nucleation between the crystals.